Wnt genes encode a large family of secreted, cysteine rich proteins that play key roles as intercellular signaling molecules in a wide variety of biological processes (for an extensive review see (Wodarz and Nusse 1998)). The first Wnt gene, mouse wnt-1, was discovered as a proto-oncogene activated by integration of mouse mammary tumor virus in mammary tumors (Nusse and Varmus 1982). Consequently, the involvement of the Wnt pathway in cancer has been largely studied. With the identification of the Drosophila polarity gene wingless (wg) as a wnt-1 homologue (Cabrera, Alonso et al. 1987; Perrimon and Mahowald 1987; Rijsewijk, Schuermann et al. 1987), it became clear that wnt genes are important developmental regulators. Thus, although at first glance dissimilar, biological processes like embryogenesis and carcinogenesis both rely on cell communication via identical signaling pathways. In a current model of the pathway, the secreted Wnt protein binds to Frizzle cell surface receptors and activates the cytoplasmic protein Dishevelled (Dsh). Dsh then transmits the signal to a complex of several proteins, including the protein kinase Shaggy/GSK3 (Sgg), the APC tumor supressor, the scaffold protein Axin and β-Catenin (β-Cat), the vertebrate homologue of Drosophila Armadillo. In this complex β-Cat is targeted for degradation after being phosphorylated by Sgg. After Wnt signaling and the resulting down-regulation of Sgg activity, β-Cat (or its Drosophila homologue Armadillo) escape from degradation and accumulate into the cytoplasm. Free cytoplasmic β-Cat translocates to the nucleus by a still obscure mechanism, and modulates gene transcription through binding the Tcf/Lef family of transcription factors (Grosschedl R 1999). Mutations of β-Cat itself or of negative regulatory elements, like APC and Axin, that lead to nuclear accumulation of β-Cat and consequently to constitutive activation of the Wnt pathway have been observed in many types of cancers, including colon, skin and breast cancer (Barker N 1999; Morin 1999; Potter 1999; Roose and Clevers 1999; Waltzer and Bienz 1999). Currently, there are no known therapeutic agents effectively inhibiting β-Cat transcriptional activation. This is partly due to the fact that many of the essential components required for its full activation and nuclear translocation are still unknown. Consequently, there is an urge to understand more about this pathway to develop effective drugs against these highly malignant diseases.
In order to identify new components required for Wingless (Wg) activation we used a Drosophila genetic approach. Specifically, we screened for dominant suppressors of the rough eye phenotype caused by a transgene that drives ectopic expression of Wg, the Drosophila homologue of Wnt, during eye development. Three genes were identified: the β-cat homologue armadillo (arm), the tcf/lef-1 homologue pangolin (pan) and legless (lgs), a completely new gene. We subsequently cloned lgs and confirmed its in vivo requirement for Wg signal transduction in embryo and in developing tissues. Epistasis experiments revealed that Lgs is at the same level or downstream of Arm. In addition, we found that the Lgs protein binds to and translocates to the nucleus with Arm in mammalian cells. Biochemical experiments confirmed the binding of Lgs to Arm. Lgs forms a tri-molecular complex with Pan and Arm and enhances the transcriptional activity of the complex.
Sequence homology search using the Blast search tool at NCBI revealed at least two human proteins sharing short amino acids domains with up to 66% sequence identity with Drosophila Lgs (dLgs). One of them, hLgs/Bcl9, has been previously implicated in B cell malignancies (Willis, Zalcberg et al. 1998). The other, hLgs-1, is a completely new gene. Several Expressed Sequence Tags (EST) could be found for both human homologues in the public human genome database, demonstrating the presence of their gene products in human normal and tumor tissues. Subsequent genetic and biochemical experiments confirmed the functional homology of hLgs to dLgs. Particularly, hLgs/Bcl9 not only binds to β-Cat and its Drosophila homologue Armadillo (Arm), but is also able to substitute for lack of dLgs during fly development. Furthermore, point mutations or deletions in the homology domains between dLgs and hLgs disrupt Lgs function, highlighting the essential role of these evolutionary conserved domains.
Lgs thus represents an exquisite target for all the diseases caused by the over-activation of the β-Cat/Tcf complex.